Effects of White and Blue-Red Light on Growth and Metabolism of Basil Grown under Microcosm Conditions

Indoor farming of basil (Ocimum basilicum L.) under artificial lighting to support year-round produce demand is an area of increasing interest. Literature data indicate that diverse light regimes differently affect downstream metabolic pathways which influence basil growth, development and metabolism. In this study, basil was grown from seedlings to fully developed plants in a microcosm, an innovative device aimed at growing plants indoor as in natural conditions. Specifically, the effects of white (W) and blue-red (BR) light under a photosynthetic photon flux density of 255 μmol m−2 s−1 on plant growth, photochemistry, soluble nutrient concentration and secondary metabolism were investigated. Plants grew taller (41.8 ± 5.0 vs. 28.4 ± 2.5 cm) and produced greater biomass (150.3 ± 24.2/14.7 ± 2.0 g vs. 116.2 ± 28.3/12.3 ± 2.5 g fresh/dry biomass) under W light compared to BR light. The two lighting conditions differently influenced the soluble nutrient concentration and the translocation rate. No photosynthetic stress was observed under the two lighting regimes, but leaves grown under W light displayed higher levels of maximum quantum yield of PSII and electron transport rate. Sharp differences in metabolic patterns under the two lighting regimes were detected with higher concentrations of phenolic compounds under the BR light.


Introduction
In recent times, indoor farming, after the initial interest in space colonization by humans [1], has met increasing interest among farmers as a practical and effective approach to support year-round produce demand in urban environments, and also to face the adverse effects of climate change on conventional agricultural systems [2][3][4][5][6][7]. So far, the development of indoor farming technology has been mainly based on progress in artificial lighting and environmental controlling/sensing, which enable farmers to grow plants under totally controlled environments. Plant lighting based on light emitting diodes (LEDs) is probably the most relevant technological progress in this field, thanks to several traits, including smaller lamp sizes, lower heat emissions, greater photonic emission efficiency, quicker reaching the stationary radiant flux, longer life span and easier connection to digital control systems compared to conventional light sources [8][9][10][11][12][13]. LED technology also

Plant Growth and Development
Basil plants grew healthy and vigorously in the two microcosms for the whole cultivation period. They reached their maximum height at the beginning of the flowering stage, i.e., at the emission of the inflorescence axes (Figure 1), which occurred 58 days after the transplant.
of all the potential of leaf and stem production is request by farmers. The effects of the two light spectra were assessed by measuring the biomass yield, as this is the first feature the farmers are interested in, and detecting the soluble nutrient and secondary metabolite contents, as these features affect the global quality of the final product. Since biomass yield and quality are both related to metabolic pathways driven by photosynthesis, the response of the photosynthetic apparatus to the different lighting conditions was also tested.

Plant Growth and Development
Basil plants grew healthy and vigorously in the two microcosms for the whole cultivation period. They reached their maximum height at the beginning of the flowering stage, i.e., at the emission of the inflorescence axes (Figure 1), which occurred 58 days after the transplant.  Figure 4. The average total fresh weights of the aerial parts under W and BR light were about 150.3 ± 24.2 g and 116.2 ± 28.3 g, respectively, whereas the corresponding dry weights were about 14.7 ± 2.0 g and 12.3 ± 2.5 g, respectively. (average height cm 41.8 ± 5.0 and 28.4 ± 2.5, respectively) ( Figure 2). The average number of leaves per plant was also significantly higher in the plants grown under the W light compared to the BR light (200. 5 ± 15.4 and 146.4 ± 30.4, respectively) ( Figure 3). The average fresh and dry weights of aerial organs are reported in Figure 4. The average total fresh weights of the aerial parts under W and BR light were about 150.3 ± 24.2 g and 116.2 ± 28.3 g, respectively, whereas the corresponding dry weights were about 14.7 ± 2.0 g and 12.3 ± 2.5 g, respectively.   (average height cm 41.8 ± 5.0 and 28.4 ± 2.5, respectively) ( Figure 2). The average number of leaves per plant was also significantly higher in the plants grown under the W light compared to the BR light (200. 5 ± 15.4 and 146.4 ± 30.4, respectively) ( Figure 3). The average fresh and dry weights of aerial organs are reported in Figure 4. The average total fresh weights of the aerial parts under W and BR light were about 150.3 ± 24.2 g and 116.2 ± 28.3 g, respectively, whereas the corresponding dry weights were about 14.7 ± 2.0 g and 12.3 ± 2.5 g, respectively.

Chlorophyll Fluorescence Analysis
The maximum photochemical efficiency (Fv/Fm) and the electron transport rate (ETR) were measured on middle fully expanded leaves 28 days after transplant, and the results are shown in Figures 5 and 6, respectively.

Chlorophyll Fluorescence Analysis
The maximum photochemical efficiency (Fv/Fm) and the electron transport rate (ETR) were measured on middle fully expanded leaves 28 days after transplant, and the results are shown in Figures 5 and 6, respectively.

Figure 4.
Average fresh (f, upper) and dry (d, lower) weights of leaves (L), stems (S) and flowers (F) and average weights of the total aerial biomasses (T) of plants grown under W and BR lights determined at the harvest stage. Bars show the mean values and the SE (n = 6 replicates). Not significant and significant differences at p ≤ 0.05, 0.01 or 0.001 are indicated as ns, *, ** and ***, respectively.

Chlorophyll Fluorescence Analysis
The maximum photochemical efficiency (Fv/Fm) and the electron transport rate (ETR) were measured on middle fully expanded leaves 28 days after transplant, and the results are shown in Figures 5 and 6, respectively.  The Fv/Fm ratio reflects the maximum quantum efficiency of PSII. The recorded mean Fv/Fm values ranged between 0.770 and 0.785, indicating no detectable photosynthetic stress in plants under both lighting regimes. As the light induction curves showed, basil grown under the W light displayed a higher electron transport rate (ETR) as compared to BR, where the maximum ETR values were reached at 997 µmol·m −2 ·s −1 of actinic light. Table 1 reports the concentrations of the soluble fractions of several main nutrients plus Na + and Clions, determined at the harvest stage in the stems and leaves of plants grown under the two lighting regimes. Table 1. Average concentrations (mg•g −1 ) ± SE (n = 3 replicates) of the soluble fractions of the main nutrients plus Na + and Clions in leaves and stems of basil plants grown under white (W) and bluered (BR) light. p values are shown; not significant and significant differences at p ≤ 0.05, 0.01 or 0.001 are reported as ns, *, ** and ***, respectively. The Fv/Fm ratio reflects the maximum quantum efficiency of PSII. The recorded mean Fv/Fm values ranged between 0.770 and 0.785, indicating no detectable photosynthetic stress in plants under both lighting regimes. As the light induction curves showed, basil grown under the W light displayed a higher electron transport rate (ETR) as compared to BR, where the maximum ETR values were reached at 997 µmol·m −2 ·s −1 of actinic light. Table 1 reports the concentrations of the soluble fractions of several main nutrients plus Na + and Cl − ions, determined at the harvest stage in the stems and leaves of plants grown under the two lighting regimes.

Metabolic Profiling
The metabolic patterns of hydroalcoholic extracts achieved from the apical (A) and middle (M) leaves grown under BR and W light obtained by LC-ESI-FT-(Orbitrap)-MS analysis in negative ionization mode are reported in Figure 7.

Metabolic Profiling
Metabolic pa erns of hydroalcoholic extracts achieved from the apical (A) and middle (M) leaves grown under BR and W light obtained by LC-ESI-FT-(Orbitrap)-MS analysis in negative ionization mode are reported in Figure 7. High-resolution mass spectrometry analysis followed by MS/MS fragmentation experiments allowed detection of 30 main compounds, mainly belonging to flavonoids, phenylpropanoids, organic acids and catechin, in addition to stilbenes and two triterpenic acids ( Table 2).  High-resolution mass spectrometry analysis followed by MS/MS fragmentation experiments allowed detection of 30 main compounds, mainly belonging to flavonoids, phenylpropanoids, organic acids and catechin, in addition to stilbenes and two triterpenic acids ( Table 2).  The LC-ESI/LTQOrbitrap/MS data were subjected to multivariate analysis using PLS-DA as a projection method, and the score scatter plot is reported in Figure 8A. The first component expressed 31% of the variance, whereas the second component accounted for 14% of the variance, allowing discrimination of the samples in two main clusters. The metabolites detected in plants grown under the W light were mainly distributed in the right part of the plot, while those observed in basil grown under the BR light were mostly spread in the left part of the plot. Due to the large number of variables, no marker compounds could be selected in the loading plot obtained from the untargeted PLS-DA analysis, as shown in Figure 8B. To identify marker metabolites useful to discriminate samples from plants grown under the two different lighting regimes, a pseudo-targeted approach was developed by building a new data matrix considering and manually measuring the areas of the peaks of the compounds listed in Table 2. This matrix was then submi ed to multivariate analy- To identify marker metabolites useful to discriminate samples from plants grown under the two different lighting regimes, a pseudo-targeted approach was developed by building a new data matrix considering and manually measuring the areas of the peaks of the compounds listed in Table 2. This matrix was then submitted to multivariate analysis by the SIMCA-P+ software using both PCA and PLS-DA. The score scatter plot of the targeted PLS-DA analysis ( Figure 9A) allowed to differentiate the samples into two groups, one from the plants grown under the W light (WA and WM), positioned in the left quadrants, and one from the plants grown under the BR light (BRA and BRM), positioned in the right quadrants, confirming the results of the untargeted analysis. An additional separation was obtained between the BRA samples that were positioned in the upper right side of the plot and the BRM samples that were placed in the lower right part. The results from the loading plot analysis ( Figure 9B) indicated that the phenolic compounds were more expressed in BRA and BRM samples.

Discussion
The experimental setup used in this study was suitable to investigate the effect of different lighting regimes on fully expanded basil plants, which represents the less investigated plant model in the field of plant-light interaction compared to smaller plants and microgreens. An experimental design based on W vs BR light under the same PPFD level was chosen as a first test of the effect of different light spectra on basil yield and quality. Biometric determinations indicated that the average dry biomass of all the aerial organs was higher in plants grown under W light. This finding disagrees with the results reported by other authors, who indicated be er growth performance in basil microgreens and in basil grown in small pots under selected wavelengths or blue-red light. Different results can be due to different basil genotypes and different lighting conditions (for example, B:R ratios, PPFD levels, photoperiod, etc.) as well as to different plant developmental stages and canopy widths [26,37,42]. In particular, the greater growth of basil under W light observed in the present study can also be related to the be er a itude of the white light to

Discussion
The experimental set-up used in this study was suitable to investigate the effect of different lighting regimes on fully expanded basil plants, which represents the less investigated plant model in the field of plant-light interaction compared to smaller plants and microgreens. An experimental design based on W vs BR light under the same PPFD level was chosen as a first test of the effect of different light spectra on basil yield and quality. Biometric determinations indicated that the average dry biomass of all the aerial organs was higher in plants grown under W light. This finding disagrees with the results reported by other authors, who indicated better growth performance in basil microgreens and in basil grown in small pots under selected wavelengths or blue-red light. Different results can be due to different basil genotypes and different lighting conditions (for example, B:R ratios, PPFD levels, photoperiod, etc.) as well as to different plant developmental stages and canopy widths [26,37,42]. In particular, the greater growth of basil under W light observed in the present study can also be related to the better attitude of the white light to reach the lower and inner parts of the plants compared to the blue and red wavelengths alone [60]. Chlorophyll fluorescence analysis was carried out on middle fully expanded leaves at about half of the expected vegetative growing period (4 weeks after transplant), in order to evaluate the effect of the lighting regimes on photosynthetic systems when plants were still in the active growing period [61]. The recorded Fv/Fm values indicated that neither W nor BR lights caused any photosynthetic stress to the plants, while the W light resulted in elevated electron transport, in line with previous results about the effect of different light spectra on PSII quantum efficiency [21,[62][63][64]. The concentrations of the soluble fractions of several main nutrients detected in the plants at the end of the growing period agree with the values already reported in basil by Yang and Kim [65]. The greater amount of the total ions in plants grown under W light, especially in stems, is consistent with the higher biomass production displayed by these plants. The soluble nutrient concentrations were influenced by the lighting conditions, except for Na + in the leaves. Markedly, the plants grown under W light accumulated higher levels of NO 3 − , K + and Cl − , and lower levels of NH 4 + , P, PO 4 3− , Mg 2+ , Ca 2+ , S, SO 4 2− . The leaf-to-stem concentration ratios were higher under the BR light for all the tested elements, except for S and SO 4 2− , thus suggesting an enhancing effect of the BR light on the translocation rate.
The importance of a metabolomics approach to evaluate the plant response to environmental factors is receiving increasing acknowledgement by the scientific literature [57]. Specifically, the metabolomics of fully expanded plants is less investigated compared to smaller plants and microgreens. To investigate the effect of the W and BR light on the secondary metabolism of the fully expanded basil plants, a metabolomics approach based on LC-ESI-Orbitrap-MS combined with a multivariate data analysis was followed. As expected, different metabolic profiles in middle and apical leaves collected from plants grown under W and BR light were recorded. Sharp differences were observed in the metabolite distribution following the multivariate analysis, confirming that the two lighting regimes differently affected the secondary metabolism in basil. In this study, an untargeted metabolomics approach was chosen, as it enables the identification of key changes in metabolic pathways and helps to reveal important and putative novel metabolites or pathways for the implementation of further analyses. Notwithstanding the great progress that has been made in this field over the past decade, plant metabolomics with an untargeted approach still seems to be a valuable approach as it can generate comprehensive information regardless of the high complexity of plant metabolites [66][67][68][69]. Among the compounds responsible for the separation of the samples obtained from plants grown under BR light, several molecules with peculiar biological and pharmacological properties were identified. In particular, antioxidant activity and beneficial effects on spermatogenesis were described for ellagic acid (Table 2, compound 7) and chlorogenic acid (Table 2, compound 9) [70], antioxidant, anti-inflammatory, antiviral and immune-stimulating properties were reported for chicoric acid ( Table 2, compound 8) [71], already detected in different basil organs [72] and antiviral, antimicrobial and anti-inflammatory activities were observed for rosmarinic acid ( Table 2, compound 13) [53]. Higher levels of these compounds were detected in plants grown under BR light, thus suggesting that the BR light can promote the production of phenolic compounds associated with interesting biological activities.

Germplasm and Growing Conditions
Seedlings of basil type Genovese cv. "Bonsai" (Blumen Vegetable Seeds, Milano, Italy), a rather common Italian variety of basil variety characterized by compact growing habit and heavily cup-shaped leaves, were transplanted in two microcosm devices, each one equipped with 6 cylindrical pots (60 cm height × 20 cm diameter; 25 cm spaced out from each other) and set up as described in d'Aquino et al. [33]. In each cylindrical pot previously filled with commercial potting soil (60% blond peat, 20% brown peat, 20% pumice 3-6 mm, pH 6.5), three seedlings were jointly transplanted. Environmental conditions were 20-26 • C (nightday) with 60% relative humidity and 18-22 • C (night-day) in the epigeal and hypogeal chambers, respectively. Each microcosm was equipped with 6 square lamps, specifically designed and supplied by Becar S.r.l. (Beghelli group), carrying different LED arrays and enabling fine regulation of light spectra and PPFD. In the 'microcosm white' (W) only LEDs Luxeon SunPlus 20 Cool White (Lumileds, Schiphol, The Netherlands) and Oslon ® SSL 80 Cool White (Osram Opto Semiconductor, Regensburg, Germany) were activated, whereas in the 'microcosm blue-red' (BR) only LEDs, Royal Blue 445-455 nm and Deep Red 655-670 nm (Lumileds) and LEDs Oslon ® SSL 80 Deep Blue 451 nm and HyperRed 660 nm (Osram Opto Semiconductor) were activated. Figure 10 reports the spectral distributions of the two lighting setups in the region λ 350 ÷ 800 nm at 87 cm distance from the light source, i.e., at the seedlings level, determined using a spectroradiometer OL-770VIS (Gooch and Housego, Ilminster, UK) equipped with an Optopolymer integrating sphere.  Figure 10 reports the spectral distributions of the two lighting setups in the region λ 350 ÷ 800 nm at 87 cm distance from the light source, i.e., at the seedlings level, determined using a spectroradiometer OL-770VIS (Gooch and Housego, Ilminster Somerset, UK) equipped with an Optopolymer integrating sphere. . Photoperiodical conditions were 16/8 h light/dark. To prevent any external effect on lighting in the epigeal chambers, appropriate shadowing of the two microcosms was implemented. The plants were watered with 8.5 l water/pot, batching the total amount along the entire cultivation period according to the biomass growth, and they were fertilized twice with Fertiactyl GZ ® (Timac) (0.5% and ammonium sulphate 1 g l −1 ) during the growing cycle. To exploit the whole vegetative growth potential of plants, the aerial parts were harvested 58 days after transplant, when all the plants were at the beginning of the flowering stage, i.e., when the emission of inflorescence axis had started and before the anthesis had occurred, to prevent hijacking of leaf and stem resources by the reproductive organs.

Biometric Determinations
Plants (n = 3) in each pot were considered as one replicate. At the harvest stage (58 days after the transplant), the plant heights and the number of leaves per plant were recorded. Leaves, stems and inflorescence axes were then collected, and fresh and dry . Photoperiodical conditions were 16/8 h light/dark. To prevent any external effect on the light supplied in the epigeal chambers, appropriate shadowing of the two microcosms was implemented. The plants were watered with 8.5 l water/pot, batching the total amount along the entire cultivation period according to the biomass growth, and fertilized twice with Fertiactyl GZ ® (Timac) (0.5% and ammonium sulphate 1 g l −1 ) during the growing cycle. To exploit the whole vegetative growth potential of plants, the aerial parts were harvested 58 days after transplant, when all the plants were at the beginning of the flowering stage, i.e., when the emission of inflorescence axis had started and before the anthesis had occurred, to prevent hijacking of leaf and stem resources by the reproductive organs.

Biometric Determinations
Plants (n = 3) in each pot were considered as one replicate. At the harvest stage (58 days after the transplant), the plant heights and the number of leaves per plant were recorded. Leaves, stems and inflorescence axes were then collected, and fresh and dry weights were determined. Samples from each pot were jointly collected. The analysis was performed on a total of 6 groups of plants per light treatment.

Chlorophyll Fluorescence Analysis
Photochemical parameters were measured on intact fully expanded leaves 28 days after transplant using a Mini version of the Imaging-PAM fluorimeter (Heinz Walz GmbH, Effeltrich, German), as described in d'Aquino et al. [33].

Ion Chromatography Analysis
Plant samples were allowed to dry in a ventilated environment at room temperature until a constant weight had been reached. Dried plant organs (stem and leaves) previously pooled from each pot were extensively grinded using a Retsch MM400 ball mill (Verder Scientific, Pedrengo, Italy). Each sample (0.5 g) was treated with 20 mL of ultrapure water in an ultrasonic bath twice for 30 min, to extract the soluble fraction; solutions were then filtered with pre-syringe filters with a porosity of 0.2 µm. One ml of each solution was then added to 100 µL of H 2 O 2 and treated for 20 min in a 705 UV Digester (Metrohm, Origgio, Italy) to digest the organic fraction. Finally, volumes were corrected to 10 mL (dilution 1:10) with ultrapure water. The concentrations of major ions were determined by ion chromatography (IC) using a Dionex ICS1100 system (Thermo Fisher Scientific, Waltham, MA, USA). The detection of NO 3 − , PO 4 3− , SO 4 2− and Cl − was performed using an ASRS 300-4 mm suppressor with a current of 33 mA, a AS22 column working with a cell volume of 100 µL and a buffer solution of 3.5 mM of sodium carbonate/bicarbonate as eluent, at a flow rate of 1.20 mL/min. The detection of Na + , K + , NH 4 + , Mg 2+ and Ca 2+ was performed using a CERS 500-4 mm suppressor with a current of 15 mA, a CS12A column working with a cell volume of 25 µL and a 20 mM methane sulfonic acid solution as eluent, at a flow rate of 0.25 mL/min. Calibration curves were calculated using certified multistandard solutions. The detection of NH 4 + , NO 3 − , PO 4 3− , SO 4 2− , Ca 2+ , Mg 2+ , Na + and Cl − was performed on dried samples, as already reported [33]. The contents of N, P and S in relation to their corresponding inorganic soluble fractions were calculated using the molar mass of the elements from the concentrations of their ionic chemical forms (NO 3 − + NH 4 + , PO 4 3− and SO 4 2− , respectively).

Plant Extraction and Hyphenated Liquid Chromatography High-Resolution Mass Spectrometry (LC-ESI-Orbitrap-MS) Analysis
Intact apical and middle leaves were independently collected from plants randomly selected in the two microcosms. Dried plant leaves were ground using liquid nitrogen and 150 mg of the powder was homogenized with 2 mL of a solution of ethanol and water (1:1). Extracts were sonicated for 10 min and centrifuged at 3000 g. Supernatants were dried under nitrogen flow and then diluted with 2 mL of methanol. To remove chlorophyll, 1 mL of extract was subject to solid phase extraction using a Strata ® SCX 55 µm, 70 Å cartridge (Phenomenex, Torrance, CA, USA) preconditioned with methanol. Elution of samples was performed using 1 mL of methanol. The eluted samples were evaporated under nitrogen flow and dissolved in methanol/water (1 mg/mL) and 10 µL was injected in the LC-MS system. LC-MS analysis and molecule identification were carried out as detailed by d'Aquino et al. [33]. For the fragmentation studies, a data-dependent scan experiment was performed to select precursor ions corresponding to the most intense peaks in LC-MS analysis. Xcalibur software version 2.1 was used for instrument control, data acquisition and data analysis.

Data Analysis
Data from biometric determination, chlorophyll fluorescence measures and ion concentrations were analysed by one-way Anova using the SPSS 27 software package (www. ibm.com/software/analytics/spss 20 November 2022). For the metabolomics analy-ses, multivariate data analysis was carried out as described by Sarais et al. [73]. Raw LC-ESI/LTQOrbitrap/MS data were analysed by MZmine software (http://mzmine. sourceforge.net/ 20 November 2022). The resulting data matrices from untargeted (48 observations and 2500 variables) and pseudo-targeted (48 observations and 30 variables) analyses were processed using Umetrics SIMCA-P+ software 12.0 using PCA (Principal Component Analysis) for visualization and PLS-DA (Partial Least Square-Discriminant Analysis) for classification.

Conclusions
In this study, higher biomass yield was recorded under W light compared to BR light in fully expanded basil plants, possibly because the W light ensures a better reach of inner and lower parts in plants in which the epigeal part displays a complex aerial architecture, with many leaf layers and extensive lateral shooting. This finding should be taken into consideration in indoor basil farming targeted to the production of fully expanded plants, in which canopy complexity increases during the cultivation period, and it is also emphasized by lateral shooting that occurs after successive harvesting of leaves. Even if a definitive relationship between metabolic profiles and basil quality is yet to be deciphered, the results from metabolomics analysis confirmed that the light spectrum affects secondary metabolism in basil and provided novel information about the metabolic profile of fully expanded basil plants. Several potentially bioactive phenolic compounds were recorded at a higher level in plants grown under BR light and this finding strengthens the hypothesis that light modulation can provide farmers with a tool to drive the basil metabolic profile according to specific aims. The two tested lighting regimes also differently affected translocation rate and leaf accumulation of minerals that are related to basil final quality, particularly nitrate ions. Overall, our results suggest that W light can be preferred in the early stages of basil cultivation to enhance biomass production, while BR light can be supplied in the late cultivation period to decrease nitrate content and to increase the content of beneficial mineral nutrients and phenolic compounds in the leaves. Nevertheless, further investigations under microcosm conditions using additional wavelengths, different PPFD levels and different photoperiodic conditions are needed to assess new lighting approaches suitable to increase the yield and to drive secondary metabolism in basil.